The importance of gene expression

Gene expression is the process by which a gene’s DNA sequence is converted into the functional proteins of the cell. Gene expression profiling is the simultaneous measurement of the cellular concentration of different messenger RNAs, often representing thousands of genes in biological samples. By comparing data from different biological specimen, differences between different types of samples or biological processes can be studied. It makes use of DNA microarrays, which contain representative sequences of the genes to be measured at one time. Labeled RNA extracted from biological samples reacts with nucleotide sequences on the microarray and the relative abundance is determined with a fluorescence scanner on each position on the chip.

Main applications

Originally, gene expression profiling was mainly applied in target discovery and validation. However, the initial promise of being able to find new targets easily was not always met. Validation of targets identified through genomics proved to be a lengthy process. Recently, the use of gene expression profiling has been expanded to later stages in the drug discovery pipeline, including pharmacogenomics-based assessment of efficacy and safety of novel compounds. The use of gene expression for clinical applications, such as patient classification and diagnostics, is emerging. Many groups are working on the classification of different types of cancer and on the development of gene expression-based diagnostic tools to select the best treatment. One such recently approved test is Mammaprint by Agendia, which allows predicting the likelihood of breast cancer returning within five to ten years after a women’s initial cancer.

Challenges

One of the largest challenges in gene expression profiling remains the establishment of good experimental design practices. Optimally, one would like to predefine the technical approach and analysis strategy to address the biological questions of the project. At Organon, we have established an expert group containing biologists, bioinformaticians and statisticians that evaluate all potential experiments before they are conducted. This has led to an increased quality in our experiments during the past few years. This task will be more difficult in studies that make use of clinical samples because it is not always possible to control the experiment as well as in a laboratory setting, like in the collection of all samples that fit the ideal design. Methods will have to be established to deal with such situations.

Another challenge is the vast bioinformatics involved with analyzing gene expression data, especially when comparing data from other “omics” technologies, such as proteomics or metabolomics. Currently there are not many tools available that easily allow combining such data, to lead the scientist to biological interpretation and planning of follow-up experiments.

Cost

If one defines the experiments well with a clear question in mind, then the outcome is mostly worth the investment. “Fishing experiments” without clear goals should be avoided and often don’t give satisfactory answers. Hence the aforementioned need for good experimental design.

Personalized medicine

Personalized medicine is based on the fact that patients differ from one to another. Recently, a multitude of genetic alterations were found in oncology, opening the possibility to stratify patients on the presence of gene mutations or over expressions. Thus, codiagnostics are being marketed that allow patients to be pre-screened for likelihood of treatment success, exemplified by Herceptin. Earlier in the R&D process, phase II and III clinical trials have failed due to genetic and nongenetic variations in the patient population. Gene expression profiling of blood cells or tissue biopsies could aid in the early assessment of profiles to pre-select patient responders and non-responders thus increasing the success rate of drug development.

Biomarkers

Gene expression profiling has developed into a robust and standardized profiling technology that allows a direct translation from comprehensive gene expression analysis to biological interpretation of a system. Its output can be translated into a multiplex gene transcript profile that can be used as a biomarker itself, or forms the basis for the validation and application of protein biomarkers. Thus, it is established firmly as a key technology in the tool kit of modern pharmaceutical and diagnostic companies.